Visual Classification

Visual classification aims to automatically categorize images, a fundamental task in computer vision with applications ranging from medical diagnosis to autonomous driving. Current research emphasizes improving generalization to unseen data and enhancing model interpretability, exploring techniques like logical reasoning regularization and incorporating language-based descriptions from large language models alongside traditional deep learning architectures (e.g., CNNs, transformers). These advancements address limitations in existing methods, such as catastrophic forgetting and reliance on spurious correlations, leading to more robust and explainable visual classification systems.

Papers